DEV Community

ABROWNFOX001
ABROWNFOX001

Posted on

Production Reality for a BTC 5m Bot

GitHub: https://github.com/abrownfox0/abrownfox001-twap60-prediction-trigger-system

YouTube walkthrough: https://www.youtube.com/watch?v=XzhugRL6BV4

Live profile: https://polymarket.com/@abrownfox001

The last series mapped what Polymarket changed.

This series is the next question:

After TWAP60, 150ms taker delay, and volume-tier limits, what does a production 5-minute bot actually have to look like?

  • Part 1 — The production contract (this post)
  • Part 2 — Slot logs that can explain a bad week
  • Part 3 — Failure modes that still kill mid-band engines
  • Part 4 — What to retune, and on which data
  • Part 5 — What I will not add to abrownfox001

The Map Is Not the Machine

Knowing the rules is cheap now.

Live 5m crypto pays 60-second Chainlink TWAP.

Crypto taker flow sits behind a 150ms delay.

CLOB order/cancel bursts sit behind volume tiers.

RTDS can go stale. Combos and agent APIs exist beside the book, not inside it.

None of that tells you whether a specific account is still a mid-band specialist or a leftover last-tick script wearing new file names.

The Production Contract I Hold the Bot To

For @abrownfox001, production means all of these at once:

1. Official 60s TWAP is the only target
2. Open pin and live score use the same series
3. Enter only in 45–55¢ when edge exists
4. Re-score inside the slot
5. Scratch decay / redeem strength
6. Assume 150ms taker delay
7. Honor rate limits
8. Hard-fail on stale official data
Enter fullscreen mode Exit fullscreen mode

If item 1 is broken, items 3–8 are decoration.

If item 6 is ignored, July fill stats are fiction.

If item 8 is skipped, the bot silently trades CEX mid again.

What Production Is Not

Production is not:

  • adding Combos to engine.ts
  • wiring Exa Connect as the live trigger
  • spraying ETH/SOL/XRP because BTC 5m got harder
  • keeping 30s features because the repo comment still says 30s
  • calling a dry-run npm start a live system

The public shell can demo the loop.

Production is the loop plus freshness, delay math, and logs.

The Only Product Still in Scope

Same specialist bet as before:

  • BTC
  • 5-minute clock
  • mid-band book
  • private P(up)
  • public execution shape in engine.ts / twapFeed.ts / clob.ts

The environment got stricter. The product did not get broader.

That is the point of this series. After a venue hardens settlement and matching, the surviving bots get narrower, not louder.

How to Audit That Claim From Outside

Do not start with the README headline.

Start with the activity tab:

  • still BTC 5m?
  • still 45–55¢?
  • still later sells than buys?
  • still same-market flat exits?
  • still not piled into the last 5 seconds?

If those hold, the production contract is at least stylistically intact.

If they break, the map changed the machine.

What Part 2 Covers

A bot that cannot explain a losing week is not in production. It is in denial.

Next post: the exact slot log I want after TWAP60 + 150ms delay — fields, timestamps, and how to separate signal decay from venue friction.

Links

Not financial advice. Paper or micro-size first.

If you have more questions, please feel free to contact me at any time: https://t.me/abrownfox001

My Polymarket Activity: https://polymarket.com/@abrownfox001?tab=activity

#Polymarket #TWAP #BTC #TradingBot #AlgoTrading #PredictionMarkets #Chainlink
Enter fullscreen mode Exit fullscreen mode

Top comments (0)